THE MORE ACTIVE MONETARY POLICY that has been pursued in the United States in recent years has resulted not only in fluctuations in the general level of interest rates but also in sharp changes in the relation between long-term and short-term rates. The spread between long-term and short-term rates was greatly narrowed during the period of rising interest rates which extended from the second half of 1954 to late 1957. Whereas for many years the yield on long-term bonds had usually exceeded rates for treasury bills and commercial paper by a considerable margin, the spread between the average yield on high-grade corporate bonds (Moody’s Aaa) and the average rate for prime commercial paper (4–6 months) diminished to only 0.05 percentage point in 1956 and 0.08 percentage point in the first ten months of 1957. With several reductions in the Federal Reserve Banks’ discount rates in November 1957 and the early months of 1958, interest rates fell rapidly. As usual, the decrease was greater for short-term rates than for long-term rates, and by June 1958 the spread between the yields on corporate bonds and commercial paper widened to 1.03 percentage points. Thereafter interest rates rose. Yields on long-term bonds soon approached their 1957 highs, but commercial paper rates remained well below their 1957 level. In the first quarter of 1959, the spread was 0.83 percentage point. The differential between yields on long-term and short-term government securities showed similar but even wider fluctuations.
During the period of almost 25 years prior to 1956–57, a substantial excess of long-term interest rates over short-term rates had come to be regarded as normal by many persons. One explanation advanced for the differential is that long-term lenders have to be offered a premium to protect them against the risk that interest rates will rise in the future, thereby causing holders of long-term securities to suffer capital losses and to miss opportunities for higher yields. Coupling this view of the usual relationship between long-term and short-term rates with the widely held theory that the long-term rate at any time reflects mainly an average of expected future short-term interest rates, J. R. Hicks concluded: “If short rates are not expected to change, the long rate will exceed the short rate by a normal risk-premium; if the current short rate is regarded as abnormally low, the long rate will be decidedly above it; the short rate can only exceed the long rate if the current short rate is regarded as abnormally high.”1 On this basis, the narrowing of the differential between long-term and short-term rates in 1956–57 might be interpreted as an indication that the market considered the high short-term rates of those years temporary. By similar reasoning, the continuance of a considerable spread when rates rose in late 1958 and early 1959 might be taken as evidence that lenders were becoming accustomed to higher interest rates and were beginning to expect them to continue.
In a longer historical perspective, however, a spread between long-term and short-term interest rates as narrow as that of 1956–57 is by no means unusual. Prior to 1930, short-term interest rates were often equal to, or higher than, long-term rates. Short-term rates fluctuated over the business cycle much more than long-term rates, usually rising above long-term rates in boom periods and falling below long-term rates at the trough of the business cycle. The difference between the periods before and after 1930 suggests the possibility that the persistent excess of long-term rates in the later period was attributable to special factors, such as the great depression, war finance, and easy money policies, and that, if more flexible monetary policies are followed in the future, the behavior of the term structure of interest rates may resemble more closely that of the period before 1930 than that of the last 25 years.
In this paper, no attempt will be made to test various hypotheses with respect to the normal relation between long-term and short-term interest rates or to account for the apparent change in that relation after about 1930. The paper is confined to a review of statistics of long-term and short-term interest rates in the United States for the past century. Some general observations on the relationship between long-term and short-term rates are presented in Section I; the cyclical variability of the relationship is examined in more detail in Section II; and the Appendix contains information on data and methods for measuring cyclical variations in interest rates.
APPENDIX: Data and Methods for Measuring Cyclical Variations in Long-Term and Short-Term Interest Rates, 1858–1958
Mr. Goode, member of the staff of the Brookings Institution, was formerly Assistant Director of the Asian Department of the Fund. Before joining the Fund staff, he was assistant professor of economics at the University of Chicago and economist at the U.8. Bureau of the Budget and the U.S. Treasury Department.
Mr. Birnbaum, economist in the Finance Division, was educated at Ohio State University and George Washington University.
J. R. Hicks, Value and Capital (Oxford, 2nd ed., 1946), p. 147. Friedrich Lutz, who also considers the long rate a reflection of expected short rates, believes that yields will be approximately equal for different maturities if no change in short rates is expected. See Friedrich A. Lutz, “The Structure of Interest Rates,” The Quarterly Journal of Economics, Vol. LV (1940–41), pp. 36–63, reprinted in American Economic Association, Readings in the Theory of Income Distribution (1946), pp. 499–529.
Frederick R. Macaulay, The Movements of Interest Rates, Bond Yields and Stock Prices in the United States Since 1856 (National Bureau of Economic Research, New York, 1938), pp. A141–61.
Data for the earlier years are from Board of Governors of the Federal Reserve System, Banking and Monetary Statistics (Washington, November 1943); those for the more recent years are from Federal Reserve Bulletins.
Rediscounting occurred among commercial banks before the establishment of the Federal Reserve System but was likely to become expensive and difficult in times of monetary stringency.
Federal Reserve Banks’ holdings of U.S. Government securities increased from $0.5 billion at the end of 1929 to $2.4 billion at the end of 1933, but remained approximately stable in the years 1934–39 The gold stock, which was $4.0 billion at the end of 1929 and again at the end of 1933, rose to $17.6 billion by the end of 1939. See Banking and Monetary Statistics (cited above), pp. 375–77.
For details of the policies of the U.S. Treasury and the Federal Reserve Board with respect to interest rates from 1939 to 1951, see Henry C. Murphy, The National Debt in War and Transition (New York, Toronto, and London, 1950), and U.S. Congress, Joint Committee on the Economic Report, Monetary Policy and the Management of the Public Debt: Replies to Questions and Other Materials for the Use of the Subcommittee on General Credit Control and Debt Management (82nd Cong., 2nd Sess., Washington, 1952), Vol. 1, pp. 50–76 and 346–68.
David Durand, Basic Yields of Corporate Bonds, 1900–1942 (National Bureau of Economic Research, Technical Paper 3, New York, 1942); David Durand and Willis J. Winn, Basic Yields of Bonds, 1926–1947: Their Measurement and Pattern (National Bureau of Economic Research, Technical Paper 6, New York, 1947); National Industrial Conference Board, The Economic Almanac, 1958 (New York, 1958), p. 85.
David Durand, op. cit., p. 4.
Ibid., pp. 12–14.
Both 1916 and 1925 were years of cyclical expansion of general business activity.
All of these curves are flat for the longer maturities (generally maturities longer than about 5 or 6 years); their shape for the shorter maturities is uncertain, but there is some indication that the curves may be descending in this range.
For 1932, the curve is flat for maturities longer than about 8 years, and un-certain for shorter maturities but possibly ascending; for 1952 and 1957, it is flat up to 10 and 20 years, respectively, and ascending thereafter.
See Chart 2 and also Wesley C. Mitchell, What Happens During Business Cycles (National Bureau of Economic Research, New York, 1951), p. 167. Mitchell presents measures of the timing and amplitude of cyclical changes of open market rates for call money, commercial paper, and 90-day time money and of yields of railroad bonds, New England municipal bonds, and high-grade corporate and municipal bonds.
See Arthur F. Burns and Wesley C. Mitchell, Measuring Business Cycles (National Bureau of Economic Research, New York, 1946).
See Appendix for additional information.
At the cyclical peak in 1873, the average call money rate is computed to be 28.51 per cent and the average commercial paper rate 15.09 per cent. The average peaks for the other 18 cycles, 1858–1933, are only 5.89 per cent and 5.77 per cent, respectively. The average peaks for the 19 cycles are 7.08 per cent and 6.26 per cent, respectively.
Arthur F. Burns and Wesley C. Mitchell, op. cit., and a private communication from the National Bureau of Economic Research to the authors of this paper.
Frederick R. Macaulay, op. cit., pp. A141–61.
Banking and Monetary Statistics (cited above), p. 451, and Federal Reserve Bulletins.
Statistics for 1920–38 from Banking and Monetary Statistics (cited above), pp. 465–67, 494, and 509–10; statistics for 1950–58 from Federal Reserve Bulletins.
Including finance company paper.
Rounded to nearest $100 million.
Frederick R. Macaulay, op. cit., pp. A141–61.
Banking and Monetary Statistics (cited above), pp. 470–71, and Federal Reserve Bulletins.
The average of the Macaulay series is 4.41 per cent in 1930 and 4.65 per cent in 1931. The average of the Moody series is 4.55 per cent in 1930 and 4.58 per cent in 1931. After 1931, the Macaulay series tends to exceed the Moody series by a wider margin.
The average of monthly bond yield quotations for the unadjusted series in 1860 is 8.591 per cent, compared with 6.040 per cent for the adjusted series, i.e., a differential of 2.551 per cent. By 1930, the average differential amounted to only 0.517 per cent. Yields indicated in the unadjusted series at all times exceed corresponding figures of the adjusted series.